Brain organoids, veganism, feminism, sometimes research & parenting life. Ruthless cosmopolitan. Critics rave: "Least annoying bad takes I've seen all week"
Cambridge, MA
- Tweets 16,913
- Following 626
- Followers 2,192
- Likes 215,241
Pinned Tweet
Because sentient animals evolved to pursue their own flourishing, not our instrumental goals, any workflow that exploits them as biological machines can in theory be made more efficient, reproducible, and humane by replacing the animals with designed-to-purpose synthetic systems.
8
15
154
56,021
People worried AI might kill everyone 15 years ago, when this was a wild hypothetical:
-weird (obviously)
-a few in a creepy or unpleasant way
People currently worried about this:
-almost everyone who knows the industry & doesn't have a gigantic COI
-also most of the ones who do
1
4
118
The former fact seems important for a small set of young vulnerable people trying to move to the Bay Area and navigate the AI safety social-institutional ecosystem.
The latter seems like, uh, the one most of us should be focusing on?
1
4
52
Dreamed Stromae released a new track about a fictional Canadian Muslim immigrant pop musician called "Ça Y Aïd". The chorus had the line "la lune, l'étoile, le show, le star/Québec, Montréal, Ottowa, Calgare"
1
7
95
fucking UNIONIZE AND EXERCISE YOUR LEGAL RIGHT TO COLLECTIVELY NEGOTIATE AND ENFORCE LEGALLY BINDING CONTRACTS WITH YOUR EMPLOYERS ALREADY omfg
"Here are some things I’m confident about. I’m confident that the people in AI labs who are issuing warnings about AI are generally sincere. I’ve talked to many people in Google Deepmind about these issues and almost all of them have sincere and serious concerns, whether or not they voice them in public."
3
9
66
5,245
Need to ban this yesterday
In 12 weeks, we built a research facility that is run entirely by AI.
AI designs, executes, and observes experiments end-to-end across biology, chemistry, and materials science.
We’re introducing SciUniverse: a benchmark that measures AI’s ability to do real-world scientific research.
4
3
29
1,544
Widespread concern yesterday that Anthropic had given autonomous AI agents a wet lab to play with despite this being obviously incredibly stupid and irresponsible, followed by fact-checks pointing out that all their benchwork was still being done by humans.
Not to worry:
In 12 weeks, we built a research facility that is run entirely by AI.
AI designs, executes, and observes experiments end-to-end across biology, chemistry, and materials science.
We’re introducing SciUniverse: a benchmark that measures AI’s ability to do real-world scientific research.
6
2
45
2,714
The steelman/motte of the "CRISPR-like" announcement is not "we made a great breakthrough" but "our models can help you inspect your raw data".
The issue with this is that they're simultaneously announcing that they're building the infrastructure to scoop you if you do that
3
11
62
1,082
Neat but not that exciting. Claude did lots of tedious sequence inspection and identified some probably-functional biology in a poorly-studied bacteriophage. There are many such systems, no evidence this one is worth the effort needed to make it into "the next CRISPR" or anything
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
3
25
209
10,002
The Q is not "is agentic AI revolutionarily good at doing tons of tool calls in service of an interesting biological idea"—this is definitely true.
It's whether they can come up with creative research ideas, then make wet experiments testing them work. This doesn't answer that
1
1
15
498
while avoiding unnecessary cruelty or disrespect towards the fellow creatures one is brought into face-to-face interactions with, and pursuing fairly quiet research into whom we should try to care for and how if society somehow, someday, overcomes morally-arbitrary speciesism
1
4
113
Proposal: EA Judaism. It's indistinguishable from regular EA, because your every action is already optimized such that following any other rule would in expectation prevent saving a nonzero number of human lives and thus violate pikuach nefesh. However you get a tax exemption
1
6
250